Putty-Clay and Investment

:
A Business Cycle Analysis
1
Simon Gilchrist
Boston University and NBER
270 Bay State Road
Boston, MA 02215
sgilchri@bu.edu
and
John C. Williams
Board of Governors of the Federal Reserve System
Mail Stop 67
Washington, DC 20551
jwilliams@frb.gov
June, 1998
Abstract
This paper develops a dynamic stochastic general equilibrium model with
putty-clay technology that incorporates embodied technology, investment irre-
versibility, and variable capacity utilization. Low short-run capital-labor sub-
stitutability native to the putty-clay framework induces the putty-clay effect of
a tight link between changes in capacity and movements in employment and
output. As a result, persistent shocks to technology or factor prices generate
business cycle dynamics absent in standard neoclassical models, including a
prolonged hump-shaped response of hours, persistence in output growth, and
positive comovement in the forecastable components of output and hours. Ca-
pacity constraints result in a nonlinear aggregate production function that im-
plies asymmetric responses to large shocks with recessions steeper and deeper
than expansions. Minimum distance estimation of a two-sector model that nests
putty-clay and neoclassical production technologies supports a significant role
for putty-clay capital in explaining business-cycle and medium-run dynamics.
Keywords: putty-clay, vintage capital, business cycle, irreversibility, capacity
utilization.
JEL Classification: D24, E22, E32
1
We are indebted to Steven Sumner for exemplary research assistance. We wish to thank Flint
Brayton, Jeff Cambell, Thomas Cooley, Russel Cooper, Sam Kortum, John Leahy, Scott Schuh,
Dan Sichel, and participants at presentations at the NBER Impulse and Propagation Workshop and
Economic Fluctuations Meeting, Columbia University, Harvard University, NYU, the University of
Maryland, the Federal Reserve Bank of New York, and the Board of Governors of the Federal
Reserve for helpful comments. The opinions expressed here are not necessarily shared by the Board
of Governors of the Federal Reserve System or its staff
1 Introduction
In this paper we develop a dynamic stochastic general equilibrium model based on
the putty-clay technology introduced by Johansen (1959). The putty-clay model
possesses a number of attractive features typically absent from models based on
neoclassical production functions, including a nonlinear short-run aggregate produc-
tion function, irreversible investment, variable capacity utilization, and endogenous
machine replacement. We investigate the implications of putty-clay technology for
macroeconomic dynamics at business cycle and medium-run frequencies.
2
A key
finding in the paper is that low short-run capital-labor substitutability native to the
putty-clay framework induces the putty-clay effect of a tight link between changes in
capacity and movements in employment and output. In the short-run, an expansion
of employment quickly confronts sharp increases in marginal costs, owing to capac-
ity constraints. Once new capacity is in place, a sustained boom in employment
and output ensues, as firms fully employ new machines without reducing utilization
rates on existing capacity.
The putty-clay technology described above has two major implications for
business-cycle dynamics. First, persistent shocks to technology or factor prices
generate a prolonged hump-shaped response of hours, persistence in output growth,
and positive comovement between the forecastable components of output and hours.
These features of the business cycle, documented by Cogley and Nason (1995) and
Rotemberg and Woodford (1996), are absent in standard neoclassical models where
the response of hours peaks upon the impact of the shock and the dynamic re-
sponse of output closely follows that of the shock.
3
Second, large shocks generate
asymmetric responses of output and hours, with recessions steeper and deeper than
expansions. This asymmetric response is consistent with the empirical evidence
documented by Neftci (1984) and others, and reflects the fact that the short-run
elasticity of output with respect to labor is decreasing in the quantity of labor em-
ployed.
The empirical relevance of putty-clay technology is confirmed by minimum dis-
tance estimation of a two-sector model that formally nests both the putty-clay and
neoclassical model within a common econometric framework. We find that the dis-
2
In a recent paper, Caballero and Hammour (1998) study medium-run issues using a putty-clay
model. See also Malinvaud (1980) and Blanchard (1997).
3
Classic examples of neoclassical models are found in the real business cycle literature of Kydland
and Prescott (1982) and Hansen (1985).
1
tance between model and data moments is minimized for an estimated putty-clay
share of total output that is on the order of 50-75%. The data overwhelmingly reject
the restriction of no role for putty-clay capital.
The putty-clay model provides an intrinsically appealing description of capital
accumulation. In this framework, capital goods embody the level of technology and
the choice of capital intensity made at the time of their creation. Ex ante, the
choice of capital intensity—the amount of capital to be used in conjunction with
one unit of labor—is based on a standard neoclassical production function. Ex post,
the production function is of the Leontief form with a zero-one utilization decision
based on the output per hour of a given piece of capital relative to the prevailing
wage rate. Over time, as the economy grows and real wages rise, older vintages
of capital become too costly to operate given their current labor requirements and
they are mothballed or scrapped.
4
Putty-clay models have a long history in both the growth (Johansen (1959),
Solow (1962), Phelps (1963), Cass and Stiglitz (1969), Sheshinski (1967) and Calvo
(1976)) and investment literatures (Bischoff (1971), Ando, Modigliani, Rasche and
Turnovsky (1974)). Due in part to computational complexities, these past litera-
tures mainly limited themselves to characterizing the long-run features of a putty-
clay economy or to partial equilibrium analysis of the investment sector. More
recently, interest in real business cycle models has spurred a revival in alternative
specifications of technology, including putty-clay.
5
Atkeson and Kehoe (1994) de-
velop a model where the energy-intensity of production is the putty-clay factor.
Their model possesses the property of a cutoff rule for utilizing capital—based on
the price of energy as opposed to the wage—but their dynamic analysis focuses on
the case where capital is always fully utilized. Cooley, Hansen and Prescott (1995)
study a model where each period physical capital is assigned to plots of land, the
supply of which is assumed to be fixed for the dynamic analysis. Although this
model features variable capacity utilization, the assumption that the land intensity
of capital can be freely changed after one period effectively cuts the dynamic link
4
The effects of technological lock-in motivate the machine replacement problem first addressed
by Johansen (1959) and Calvo (1976), and more recently formalized in a dynamic programming
environment by Cooper and Haltiwanger (1993) and Cooley, Greenwood and Yorukoglu (1994).
5
Vintage models have also experienced a resurgence of late as witnessed by Benhabib and Rus-
tichini (1991), Benhabib and Rustichini (1993), Caballero and Hammour (1996), Campbell (1994),
Cooper, Haltiwanger and Power (1995), Boucekkine, Germain and Licandro (1997), and Green-
wood, Hercowitz and Krusell (1997).
2
between capital and labor that is key to the putty-clay effect.
Our model incorporates what we view as the essential features of the putty-
clay framework, including variable capacity utilization and investment irreversibil-
ity. Although the assumption of ex post Leontief technology may at first seem
unrealistically stark, the resulting aggregate production function embeds, depend-
ing on the model’s parameterization, both the relatively flat short-run supply curve
usually associated with a neoclassical model and a reverse L-shaped supply curve
traditionally associated with the putty-clay framework. In addition, the model’s
micro-foundations are largely consistent with microeconomic evidence on the impor-
tance of plant shutdowns as a short-run adjustment margin (Bresnahan and Ramey
(1994)) and the lumpiness of investment at the plant level (Doms and Dunne (1993),
Cooper et al. (1995), Caballero, Engel and Haltiwanger (1995)).
The distinguishing features of the model developed in this paper are nicely illus-
trated by the experiment of a reduction in the cost of producing new capital goods.
This reduction in capital cost raises the return to new capital and causes a surge
in investment, which over time leads to rising aggregate output and consumption.
Initially, however, firms’ efforts to raise employment encounter capacity constraints
owing to the putty-clay nature of capital. As a result of this low short-run sub-
stitutibility of labor for capital, the initial aggregate response of both output and
hours is muted. To efficiently increase production, firms invest in new capacity,
that, once in place, can be utilized by an expanded workforce. Ex post fixity of the
capital-labor ratio for existing capacity implies that labor can only be reallocated
to new machines at the cost of mothballing existing capacity. Therefore, there is
an incentive to simultaneously utilize both new and exisiting capital. This dynamic
linkage between capital and labor causes hours and output to rise together for a
sustained period of time following the initial burst of investment, generating the
putty-clay effect.
The dynamic response to a large increase in the cost of new capital goods differs
in some respects from that described above. In this case, a large fraction of the over-
all adjustment of output and hours is accomplished through an immediate reduction
in capacity utilization. Thus, the putty-clay model naturally delivers asymmetric
responses to positive and negative shocks, with the asymmetries increasing in the
magnitude of the shock.
3
2 The Model
In this section we describe the model and derive the equilibrium conditions. Each
capital good possesses two defining qualities: its level of embodied technology and
its capital intensity. The underlying or ex ante production technology is assumed
to be Cobb-Douglas with constant returns to scale, but for capital goods in place,
production possibilities take the Leontief form: there is no ex post substitutabil-
ity of capital and labor. In addition to aggregate technological change, we allow
for the existence of idiosyncratic uncertainty regarding the productivity of invest-
ment projects. As in Campbell (1994), the introduction of heterogeneity within
vintages smooths the aggregate allocation and simplifies computation of the equi-
librium. More importantly, such idiosyncratic uncertainty implies the existence of
a well-defined aggregate production function despite the Leontief nature of the mi-
croeconomic utilization choice.
Once in place, capital goods are irreversible, that is, they cannot be converted
into consumption goods or capital goods with different embodied characteristics,
and have zero scrap value. Firms can choose, however, whether or not to operate
a given unit of capital depending on the profitability of doing so in the current
economic environment. We assume that there are no costs of taking machines or
workers on- and off-line. As such, the utilization choice is purely atemporal. The
optimal utilization choice for each unit of capital is determined by the difference
between the (labor) productivity of the capital and the cost of utilizing the capital,
which in the absence of other costs equals the wage rate. If the productivity of a
unit of capital exceeds the wage rate, the capital is used in production, otherwise,
it is not. In equilibrium, the wage rate, capacity utilization rate, and levels of em-
ployment, production, consumption, and investment are determined jointly by the
dynamic optimizing behavior of households and firms. To characterize the equilib-
rium allocation, we first discuss the optimization problem at the project level and
then describe aggregation from the project level to the aggregate allocation.
2.1 The Investment Decision
Each period a set of new investment “projects” becomes available. Constant returns
to scale implies an indeterminacy of scale at the level of projects, so without loss of
generality, we normalize all projects to employ one unit of labor at full capacity. We
4
refer to these projects as “machines.” Capital goods require one period for initial
installation and then are productive for M ≥ 1 periods. The productive efficiency
of machine i initiated at time t is affected by two stochastic productivity terms,
one idiosyncratic, one aggregate. In addition, we assume all machines, regardless of
their relative efficiency, fail at an exogenously given rate that varies by the age of
the machine. In summary, capital goods are heterogeneous and are characterized by
three attributes: vintage (age and level of aggregate embodied technology), capital-
intensity, and the realized value of the idiosyncratic productivity term.
The productivity of each machine, initiated at time t, differs according to the
log-normally distributed random variable, θ
i,t
, where
log θ
i,t
∼ N(log θ
t

1
2
σ
2
, σ
2
).
The aggregate index θ
t
measures the mean level of embodied technology of vintage t
capital goods and σ
2
is the variance of the idiosyncratic shock. The mean correction
term −
1
2
σ
2
implies E(θ
i,t

t
) = θ
t
. We assume θ
t
follows a stochastic process with
mean gross growth rate (1+g)
1−α
. For the sake of notational clarity, in the following
discussion we abstract from disembodied aggregate technological change of the type
typical in the real business cycle literature. The inclusion of a stochastic disembodied
technology process is straightforward and used in section 3 when analyzing model
dynamics.
Before investment decisions are made, the economy-wide level of vintage tech-
nology θ
t
is observed but the idiosyncratic shock to individual machines is not. We
also assume that after the revelation of the idiosyncratic shock, further investments
in existing machines are not possible. Subject to the constraint that labor employed,
L
i,t+j
, is nonnegative and less than or equal to unity (capacity), final goods output
produced in period t +j by machine i of vintage t is
Y
i,t+j
= θ
i,t
k
α
i,t
L
i,t+j
,
where k
i,t
is the capital-labor ratio chosen at the time of installation. Denote the
labor productivity of a machine by
X
i,t
≡ θ
i,t
k
α
i,t
.
The only variable cost to operating a machine is the wage rate, W
t
. Idle ma-
chines incur no variable costs and have the same capital costs as operating machines.
6
6
The model can be extended to allow for a fixed labor cost per unit of capital. Under such a
5
Figure 1: Steady-state Distribution of Labor Productivity
0
1
2
3
4
0 1 2 3 4 5 6




0
1
2
3
4
0 1 2 3 4 5 6



0.0
0.1
0.2
0.3
0.4
0.5
1 2 3 4 5 6



Aggregate (left axis)
Newest Vintage(right axis)
Project level Productivity (x)
D
i
s
t
r
i
b
u
t
i
o
n

w
e
i
g
h
t
e
d

b
y

i
n
v
e
s
t
m
e
n
t
Equilibrium
Wage
Given the Leontief structure of production, these assumptions imply a cutoff value
for the minimum efficiency level of machines used in production: those with pro-
ductivity X
i,t
≥ W
t
are run at capacity, while those less productive are left idle.
To illustrate these ideas, Figure 1 shows the steady-state distribution of labor
productivity for the model calibrated to parameters specified below. The height of
the distribution reflects the number of machines at any given productivity level. The
cutoff value for the wage is shown as a vertical line. Capital goods with productivity
lying to the right of the cutoff are used in production, those to the left are idle.
Capital utilization is given by the area in the shaded region divided by the total
area under the distribution.
Figure 1 also shows the distribution of labor productivity for the most recent
vintage (right scale). Its position on the horizonal axis reflects both the current level
of technology and the capital intensity of new machines. Owing to trend growth
specification, it is optimal to permanently scrap machines whose efficiency falls below some cutoff.
This modification substantially complicates the investment decision and is left for future research.
6
and relatively long-lived capital, the average labor productivity of the most recent
vintage is substantially higher than the average labor productivity of existing ma-
chines. Obsolescence through embodied technical change implies that old vintages
have lower average utilization rates than new vintages. Note that trend growth in
investment—due to population growth and technological change—causes the aggre-
gate distribution to be skewed.
To derive the equilibrium allocation of labor, capital intensity, and investment,
we begin by analyzing the investment and utilization decision for a single machine.
Define the time t discount rate for time t +j income by
˜
R
t,t+j

¸
j
s=1
R
−1
t+s
, where
R
t+s
is the one period gross interest rate at time t + s. At the machine level,
capital intensity is chosen to maximize the present discounted value of profits to the
machine:
max
k
i,t
,{L
i,t+j
}
M
j=1
E
t

−k
i,t
+
M
¸
j=1
˜
R
t,t+j
(1 −δ
j
)(X
i,t
−W
t+j
)L
i,t+j
¸
, (1)
s.t. 0 ≤ L
i,t+j
≤ 1, j = 1, . . . , M,
0 < k
i,t
< ∞,
where δ
j
is the probability a machine has exogenously failed by j periods and ex-
pectations are taken over labor productivity, whose realization depends on the time
t idiosyncratic shock, and future values of wages and interest rates.
Because investment projects are identical ex ante, the optimal choice of the
capital-labor ratio is equal across all machines in a vintage; that is, k
i,t
= k
t
, ∀i.
Denote the average productivity of the entire stock of vintage t capital by X
t
=
θ
t
k
α
t
. Capital utilization for vintage s at time t is the ratio of labor employed to
employment capacity of the vintage, given by Pr(X
i,s
> W
t
|W
t
, θ
t
). Given the
log-normal distribution for θ
i,t
we obtain:
Pr(X
i,s
> W
t
|W
t
, θ
t
) = 1 −Φ(z
s
t
),
where Φ(·) is the c.d.f. of the standard normal and
z
s
t

1
σ

log W
t
−log X
s
+
1
2
σ
2

,
Similarly, capacity utilization for vintage s at time t is the ratio of actual output
produced from the capital of a given vintage to the level of output that could be
7
produced at full capital utilization. Letting F(·) denote the cdf of X
i,t
, capacity
utilization is formally defined as


X
is
>Wt
X
i,s
dF(X
i,s
)


0
X
i,s
dF(X
i,s
)
= (1 −Φ(z
s
t
−σ))
where the equality follows from the log-normality of X
i,t
.
7
If all machines were fully utilized, labor productivity would simply equal X
t
.
With partial utilization, labor productivity also depends on capital and capacity
utilization. The average product of labor for vintage s capital at time t, is
APL
s
t
=
1 −Φ(z
s
t
−σ)
1 −Φ(z
s
t
)
X
s
.
Letting φ(·) denote the p.d.f. of the standard normal, the marginal product of labor
for vintage s capital at time t is
MPL
s
t
=
φ(z
s
t
−σ)
φ(z
s
t
)
X
s
.
For any given vintage, the marginal product of labor is equal to the efficiency of the
least productive machine of the vintage in operation.
8
Expected net income in period t from a vintage s machine, π
s
t
, conditional on
W
t
, is given by
π
s
t
= (1 −δ
t−s
)

1 −Φ(z
s
t
−σ)

X
s

1 −Φ(z
s
t
)

W
t

.
Substituting this expression for net income into equation 1 eliminates the future
choices of labor from the investment problem. The remaining choice variable is k
t
.
The first order condition for an interior solution for k
t
is given by
9
k
t
= αE
t

M
¸
j=1
˜
R
t,t+j
(1 −δ
j
)

1 −Φ(z
t
t+j
−σ)

X
t
¸
. (2)
7
Capacity utilization may be expressed as [E(Xi,s|Xi,s > Wt)/E(Xi,t)] Pr(Xi,s > Wt). We
then use the formula for the expectation of a truncated log-normal: If log(µ) ∼ N(ζ, σ
2
), then
E(µ|µ > χ) =
(1−Φ(γ−σ))
(1−Φ(γ))
E(µ) where γ = (log(χ) −ζ)/σ (Johnson, Kotz and Balakrishnan 1994).
8
Normalizing the quantity of machines at unity, marginal product equals the increment to ma-
chine output obtained by reducing the efficiency cutoff Wt, divided by the increment to labor input
obtained by reducing the cutoff Wt. The increment to output equals
∂(1−Φ(z
s
t
−σ)Xs)
∂z
s
t
∂z
s
t
∂Wt
, while the
increment to labor equals
∂(1−Φ(z
s
t
))
∂z
s
t
∂z
s
t
∂Wt
.
9
This first order condition is obtained by taking the derivative of the profit function with respect
to kt, recognizing that in equilibrium, the marginal machine earns zero quasi-rents so that
∂π
s
t
∂z
s
t

1
σ
φ(z
s
t
−σ)Xs −
1
σ
φ(z
s
t
)Wt = 0.
8
New machines are put into place until the value of a new machine (the present
discounted value of net income) is equal to the cost of a machine (k
t
)
k
t
= E
t

M
¸
j=1
˜
R
t,t+j
(1 −δ
j
)

1 −Φ(z
t
t+j
−σ)

X
t
(3)

M
¸
j=1
˜
R
t,t+j
(1 −δ
j
)

1 −Φ(z
t
t+j
)

W
t+j
¸
This is the free-entry or zero-profit condition. The first term on the right hand
side of equation 3 reflects the expected present discounted value of output adjusted
for the probability that the machine’s idiosyncratic productivity draw is too low to
profitably operate the machine in period t + j. The second term likewise reflects
the expected present value of the wage bill, adjusted for the probability of such a
shutdown. This condition must hold as long as there is gross investment in period
t.
10
Equations 2 and 3 jointly imply that, in equilibrium, the expected present value
of the wage bill equals (1 −α) times the expected present value of revenues.
2.2 Aggregation
Aggregation of machine-level labor inputs and output is a two-step process. First,
labor input and output of machines in each vintage are aggregated into vintage
totals. Second, inputs and outputs from the M productive vintages are summed to
yield aggregate values. Total labor employment, L
t
, is
L
t
=
M
¸
j=1

1 −Φ(z
t−j
t
)

(1 −δ
j
)Q
t−j
. (4)
where Q
t−j
is the quantity of new machines started in period t−j, Φ(z
t−j
t
) is the idle
rate of those machines in period t, and δ
j
reflects the fact that a subset of machines
has failed completely. Aggregate final output, Y
t
is
Y
t
=
M
¸
j=1

1 −Φ(z
t−j
t
−σ)

(1 −δ
j
)Q
t−j
X
t−j
. (5)
In the absence of government spending or other uses of output, aggregate consump-
tion, C
t
, satisfies
C
t
= Y
t
−k
t
Q
t
, (6)
where k
t
Q
t
is gross investment in new capital machines.
10
If the cost of a machine exceeds the value of a machine for all admissible values of kt no
investment is undertaken.
9
2.3 Preferences
To close the model, we specify the economic relationships that determine labor sup-
ply and savings decisions. We assume that the economy is made up of representative
households whose preferences are given by
1
1 −γ
E
t

¸
s=0
β
s

C
t+s
(N
t+s
−L
t+s
)
ψ
N
t+s

1−γ
, (7)
where β ∈ (0, 1), γ > 0, ψ > 0, and N
t
= N
0
(1 +n)
t
is the household’s growing time
endowment.
11
Households optimize over these preferences subject to the standard
intertemporal budget constraint. We assume that claims on the profit streams of
individual machines are traded; in equilibrium, households own a diversified portfolio
of all such claims.
The first-order condition with respect to consumption is given by
U
c,t
=
β
1 +n
E
t
R
t,t+1
U
c,t+1
, (8)
where U
c,t+s
denotes the marginal utility of consumption. The first-order condition
with respect to leisure and work is given by
U
c,t
W
t
+U
L,t
= 0, (9)
where U
L,t
denotes the marginal utility associated with an incremental increase in
work (decrease in leisure). This completes our description of the economy.
The rational expectations equilibrium is defined to be the set of sequences of
prices and quantities such that each household and firm solves its respective maxi-
mization problem as described above, taking prices as given, and all markets clear.
The derivation of the deterministic steady state and its properties are found in
Gilchrist and Williams (1998b). In conducting the dynamic analysis we focus on de-
viations from the balanced growth path. The solution methodology for the dynamic
analysis is described in the appendix.
3 Model Dynamics
In this section we describe the model’s implications for business cycle and medium-
run dynamics in response to persistent shocks to factor prices; in the next section
11
We also considered the case of indivisible labor as in Hansen (1985) and Rogerson (1988). This
specification alters the magnitude of the dynamic responses reported below but not the qualitative
properties of the model.
10
we turn our attention to permanent technology shocks. The purpose of this analysis
is not to argue for a specific theory of business cycles based on particular shocks,
but instead to document the putty-clay model’s dynamic properties and their corre-
spondence to those evident in the data for a wide variety of shocks. For the purpose
of comparison, we construct a neoclassical model of vintage capital, initially intro-
duced by Solow (1962). Details of this model are provided in the appendix. In the
vintage model, the restriction that ex post capital-labor ratios are fixed is removed.
Thus, this model takes the standard Cobb-Douglas putty-putty formulation. The
two models are otherwise identical.
We focus primarily on three key business cycle properties: comovement, per-
sistence, and asymmetries. First, as emphasized by Lucas (1977), a fundamental
feature of the business cycle is that output movements across sectors exhibit positive
comovement. The real business cycle literature tends to focus on a particular inter-
pretation of comovement that is based almost exclusively on a model’s implications
for unconditional second moments of filtered data; see, for example, Kydland and
Prescott (1991). Rotemberg and Woodford (1996) extend the notion of comovement
to the forecastable components of output, hours, and consumption. Using a VAR
model, they document that the forecastable components of these variables also ex-
hibit strongly positive comovement. Second, Cogley and Nason (1995) document
that output growth displays significant positive serial correlation; that is, output
growth is persistent. Finally, there is a wide range of evidence that asymmetries
exist with respect to the business cycle and the dynamic response to particular
shocks. As documented below, the putty-clay model possesses a powerful internal
propagation mechanism that yields dramatic improvements over the neoclassical
model in all three dimensions.
3.1 The Short-Run Aggregate Supply Curve
Insight into the dynamic responses of the putty-clay model is provided by the short-
run aggregate supply curve (the inverse of the marginal product of labor). Here,
short-run refers to the time period during which the capital stock is fixed. Figure 2
shows the short-run aggregate supply curve, computed as the markup 1/W, in log
deviations from the deterministic steady state, for the neoclassical vintage model
and three versions of the putty-clay model with different degrees of idiosyncratic
uncertainty (measured by σ). For the neoclassical vintage model, the production
11
Figure 2: Short-run Aggregate Supply Curve
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
-0.30 -0.25 -0.20 -0.15 -0.10 -0.05 0.00 0.05 0.10



-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
-0.30 -0.25 -0.20 -0.15 -0.10 -0.05 0.00 0.05 0.10



-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
-0.25 -0.20 -0.15 -0.10 -0.05 0.00 0.05



-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
-0.30 -0.25 -0.20 -0.15 -0.10 -0.05 0.00 0.05



Cobb-Douglas
σ = .25
σ = .15
σ = .05
Output (log deviation from steady state)
M
a
r
k
u
p

(
l
o
g

d
e
v
i
a
t
i
o
n

f
r
o
m

s
t
e
a
d
y

s
t
a
t
e
)
function is Cobb-Douglas implying that the SRAS is linear in logs. The SRAS of
the putty-clay model, however, is distinctly nonlinear. For very low values of σ, the
putty-clay SRAS curve becomes vertical for levels of output a few percent above
steady state. As σ increases, the SRAS curve of the putty-clay model becomes
less sharply curved and approaches that of the neoclassical vintage model as σ
approaches infinity. Thus, the model developed in this paper embeds both the
reverse-L shaped aggregate supply curve traditionally associated with putty-clay
technology and the log-linear aggregate supply curve of the neoclassical production
function. The degree of idiosyncratic uncertainty determines the extent to which the
model’s short-run aggregate production function and dynamic responses are more
putty-clay or neoclassical in flavor.
In the putty-clay model, the variable slope of the SRAS curve results from vary-
ing utilization rates of existing machines. Variable utilization is often suggested as
an explanation for the fact that empirical estimates of the short-run elasticity of pro-
duction with respect to labor inputs,
d lnY
d lnL
, are much closer to unity than to labor’s
12
share.
12
The intuition here is that a 1% increase in labor effectively causes a 1%
increase in capital, through increased utilization, and hence a 1% increase in output.
This simple calculation relies on the assumption that capital goods are homogenous
however. It also ignores equilibrium determinants of utilization and capacity. In
the putty-clay model,
dY
dL
= w where w is the efficiency of the marginal machine.
As labor inputs increase, the quality of the marginal machine falls, guaranteeing
d ln(Y )
d ln(L)
< 1.
By considering optimal capacity choice, we can explicitly quantify
d ln(Y )
d ln(L)
. In
steady state, the short-run elasticity of output with respect to labor for the putty-
clay model equals 1 − α, the long-run labor share. If utilization rates rise above
steady-state, costs increase rapidly and
d lnY
d lnL
< 1 − α. The only way to justify
d lnY
d lnL
> 1 −α is to argue that firms frequently hold costly excess capacity. This is
sub-optimal from the firm’s point of view however. Hence, except in response to
large negative shocks, firms typically operate in a region where
d lnY
d lnL
1 − α, and
variable utilization does not provide measured short-run increasing returns to labor
in the putty-clay model.
Besides having important implications for utilization rates and their influence
on labor productivity, the variable slope of the putty-clay SRAS also implies that
dynamic responses to positive shocks differs from those to negative shocks, as dis-
cussed below. Although not examined here, the nonlinear aggregate supply curve
also implies asymmetries in price adjustment in models with nominal rigidities.
3.2 Calibration
The models are calibrated using parameter values taken from Christiano and Eichen-
baum (1992) and Kydland and Prescott (1991), except for the trend growth rates,
which are averages over 1954–96.
13
We assume a period is one quarter of one year.
In annual basis terms, the calibrated parameters are β = 0.97, ρ = 1, ψ = 3, g =
0.018, n = 0.015, δ = 0.084, α = 0.36, M = ∞. The results reported in this
paper are not sensitive to reasonable variations in these parameters. When cali-
brating the model, the only parameter for which we do not have a prior estimate
is the variance of the idiosyncratic component of a machine’s productivity, σ
2
. As
12
Basu and Fernald (1997) provide a recent discussion.
13
These estimates are obtained from long-run restrictions, and, with one caveat, are therefore
valid for both the putty-clay model and the neoclassical model. The caveat is that variation in
σ has a small impact on steady-state properties through endogenous depreciation. Endogenous
depreciation alters the estimated δ by 1-2% and has only a very minor effect on model properties.
13
discussed above, for large σ, the short-run aggregate supply curve is very close to
Cobb-Douglas. By lowering σ we increase the curvature of the short-run aggregate
supply curve and increase the degree to which the model displays dynamics unique
to the putty-clay structure. To make clear distinctions between the neoclassical and
putty-clay models we set σ = 0.15. Lowering σ to 0.1 does not alter model results
in any substantial manner; lowering σ much further causes numerical problems for
the dynamic solution methods. On the other hand, raising σ to 0.5 for almost all
essential purposes replicates the neoclassical model dynamics, while intermediate
values (σ = 0.2 − 0.25) provide results that are a combination of the neoclassical
and putty-clay model with low σ.
3.3 Capital Cost Shocks
We start by characterizing the effect of a temporary but persistent shock to the
cost of producing capital goods relative to consumption goods. This is identical to
an increase in technology embodied in capital goods; henceforth, we describe it as
such.
14
We assume that embodied technology follows the process (1 −ρ
θ
L) ln θ
t
=
(1 −ρ
θ
L)t ln(1 +g) +u
t
, where u
t
is an i.i.d. innovation and L is the lag operator.
We set the autocorrelation coefficient of the shock process at ρ
θ
= 0.95 implying a
half life for the shock of just under 14 quarters.
Figure 3 shows the impulse response function to output for both the putty-clay
model (upper panel) and the neoclassical model (lower panel) to a one percentage
point positive shock to embodied technology. The difference in output dynamics
between the two models is striking. For the putty-clay model, output rises very
little initially, steadily increases for a period of five years, and eventually returns to
steady state. For the neoclassical model, the peak response occurs at the onset of
the shock, after which output exhibits something close to exponential decay as it
returns to steady state. In the neoclassical model, output dynamics simply mirror
shock dynamics with no evidence of any interesting cyclical pattern. In the putty-
clay model, output exhibits a clear “hump-shaped” response that creates a slowly
unfolding and long-lasting business cycle.
Figure 3 also plots the impulse response of the Solow residual, conventionally
measured, for the putty-clay and neoclassical models. In both models, movements in
14
It is, not, however, the same as a distortionary shock to the cost of capital goods from a change
in taxes or that might result from a monetary disturbance in a model with nominal rigidities. Such
shocks affect the equilibrium allocation but not the feasible allocation for the economy.
14
Figure 3: Persistent Cost of Capital Shock
0.0
0.2
0.4
0.6
0.8
0 5 10 15 20

Putty-Clay Model

Output
Solow Residual
Adjusted Solow Residual
years
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 5 10 15 20

Neoclassical Model

Output
Solow Residual
years
Figure 3: Persistent Cost of Capital Shock
the Solow residual are much smaller (by a factor of 3 to 5) than movements in output.
Hence, in both models, embodied technological change provides “magnification” as
usually measured by movements in output vis-a-vis the Solow residual. In addition
to magnification, the putty-clay model provides significant positive comovement,
in levels and growth rates, between output and the Solow residual over the cycle.
This positive comovement stems from the fact that both output and the Solow
residual exhibit similar hump-shaped responses to embodied technological change.
In the neoclassical model, embodied shocks generate a negative correlation between
growth in output and the Solow residual over the first five years.
The third line in the top panel of figure 3 shows the Solow residual, after cor-
recting for capital utilization. Initially, consistent with the arguments made above,
15
Figure 4: Persistent Cost of Capital Shock
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
0 5 10 15 20

Investment

years
0.0
0.1
0.2
0.3
0.4
0.5
0 5 10 15 20

Labor

years
0
1
2
3
4
5
0 5 10 15 20

Q

years
-0.5
0.0
0.5
1.0
0 5 10 15 20

k

years
Putty-Clay Model
the correction has only a trivial effect on the Solow residual. Over time, we see a
somewhat larger adjustment, but this adjustment reflects expanded capacity that
lowers the utilization rate of existing machines in later periods.
The upper panels of Figure 4 show the responses of investment and labor hours
in the putty-clay model to the same shock. Investment rises immediately as the
economy seeks to build new capital goods that embody the latest technology. Hours
increase and consumption falls in response to high real interest rates. The initial
expansion in hours is muted, however, owing to the sharply increasing short-run ag-
gregate supply curve embedded in the putty-clay model. As more capital is brought
16
on line, the short-run supply curve shifts out and labor expands further. As a result,
the peak labor response occurs three years after the onset of the shock.
The slow but sustained rise in hours above steady-state levels occurs because the
benefits to building new capital goods are much greater if existing, efficient capital
is not scrapped in the process. With ex post Leontief technology, labor cannot
be reallocated across machines to equate marginal products. To benefit from new
machines without losing the productive services of existing capital, the economy
must hire new workers to operate these machines. Hence, as new machines become
operative, more labor is hired. Eventually, as the productive value of these machines
falls, labor returns to steady state. We refer to this dynamic linkage between labor
and machines as the putty-clay effect.
Decomposing the investment dynamics into the quantity of new machines, Q,
and the capital intensity of each new machine, k, adds further insight into the
model’s dynamics. As shown in the lower panels of Figure 4, at the onset of the
shock, Q rises and k falls as a large number of low capital-intensity machines are
produced. This rapid expansion of inexpensive (in terms of foregone consumption)
machines shifts the short-run aggregate supply curve to the right and facilitates
the increase in labor input. This reliance on low efficiency capital does not persist,
however. As the real interest rate falls and the real wage rises, firms substitute into
high capital-intensity capital goods.
The capital-intensity of new machines remains at elevated levels for a number
of years as households store the benefits of the temporary shock through increased
saving implying capital deepening. Rising capital intensities dramatically offset
the exponential rate of decay in technology and consequently provide a sustained
increase in the efficiency levels of new machines for a number of years after the
shock has occurred.
15
Because machines are long-lived, this sustained increase in
efficiency levels of new machines translates into a sustained increase in total labor
productivity over a long horizon.
15
In the initial period, θt is 1% above steady state while k is 1.3% below steady state, implying
that x = θk
α
, the mean efficiency of new machines, is 0.75% percent above steady state. Six years
later θ is 0.3% above steady state while k is 1.1% above steady state, implying that x is still 0.7%
above steady state.
17
3.4 Labor Cost Shocks
We now consider the effect of a temporary but highly persistent shock to the
marginal cost of labor. This shock may be interpreted as either a reduction in
the tax on wage income or a shock to preferences that reduces the marginal utility
of leisure relative to the marginal utility of consumption. Formally, we specify the
labor cost shock ln η
t
= ρ
η
ln η
t−1
+ e
t
, where e
t
is an i.i.d. innovation, and set
ρ
η
= 0.98. We embed η
t
in the labor-leisure first-order condition given by equa-
tion 9. Figure 5 shows the responses of output, hours, and consumption to a one
percentage point reduction in the marginal cost of labor for the two models. The
hump-shaped response of output and hours in the putty-clay model observed in
response to the capital cost shock carries over to the labor cost shock.
The labor cost shock illustrates the ability of the two models to generate co-
movement between output, hours, and consumption. As seen in the figure, this
comovement is especially strong in the putty-clay model after the initial shock pe-
riod and is therefore present in the forecastable components. That is, starting from
the first period, output, hours and consumption are rising for a number of years,
after which time they decline in unision. The comovement is much weaker in the
neoclassical model, especially during the first five years following the onset of the
shock. Moreover, what little positive comovement does occur during this period is
mostly unforecastable. While consumption is rising over several years, hours and
output are falling, making the forecastable comovement between consumption and
the other two series negative rather than positive.
3.5 Nonlinear Dynamics
The degree of curvature embedded in the short-run aggregate supply curve plays an
important role in conditioning the putty-clay model’s response to shocks. For small
shocks such as those discussed above, the magnitude of the response to positive
and negative shocks is roughly the same. As we consider larger disturbances, the
curvature of the short-run aggregate supply curve away from steady-state becomes
important, and the model’s dynamic responses display interesting asymmetries. In
particular, the model delivers the result that a response to negative shocks is more
rapid and larger than that to positive shocks. This is consistent with the evidence on
the response to monetary shocks documented by Cover (1992) and that to oil price
shocks studied by Tatom (1988) and Mork (1989). In a business cycle context, this
18
Figure 5: Persistent Cost of Labor Shock
0.0
0.2
0.4
0.6
0.8
0 5 10 15 20 25 30

Putty-Clay Model

Output
Hours
Consumption
years
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 5 10 15 20 25 30

Neoclassical Model

Output
Hours
Consumption
years
pattern of asymmetric responses can imply that recessions are deeper and steeper
than expansions, a result consistent with the time series evidence documented by
Neftci (1984), Sichel (1993), and Potter (1995).
For the type of shocks considered here, the most likely source of large shocks in
an economy such as the U.S. come through changes in tax policy that affect factor
prices.
16
An investment tax credit or a revision in personal income tax rates are
plausible sources of persistent movements in factor costs of 10% or more. To display
the model’s ability to generate asymmetries, we therefore consider the differential
16
Other potential sources of large shocks are energy prices and monetary disturbances.
19
Figure 6: Asymmetries in Response to Labor Cost Shocks

0
1
2
3
4
5
0 2 4 6 8 10

Labor

years
0
1
2
3
4
5
0 2 4 6 8 10

Output

years
-2
-1
0
1
2
3
4
5
0 2 4 6 8 10

Consumption

years
0
2
4
6
8
10
12
14
16
0 2 4 6 8 10

Investment

years
10% Increase
10% Decrease
* Impulse Responses to 10% Increase in Cost of Labor Displayed with Reverse Sign
effect of a 10% increase versus decrease in the labor cost shock considered above.
The results reported in Figure 6 reveal the basic source of the asymmetry. A
labor cost increase causes an immediate shutdown of machines as the economy
moves down the relatively flat portion of the short-run aggregate supply curve.
This immediate shutdown produces a sharp contraction in output and hours in
both the initial and subsequent periods. Owing to the shutdown, the economy has
excess machine capacity, and investment drops sharply in response to the shock. In
contrast, a labor cost decrease has little immediate effect on either output or hours as
the economy is pushed up the steep portion of the short-run aggregate supply curve.
20
Evidence of non-linearity only disappears after 6-8 years as capacity eventually
adjusts. As a result, in the putty-clay model, large contractionary shocks cause steep
immediate declines in output and hours while large expansionary shocks generate
a hump-shaped dynamic response even more pronounced than in the case of small
expansionary shocks. Of particular interest here is the fact that the asymmetries
on labor are the most pronounced, a result supported by Neftci (1984)’s non-linear
time series analysis.
4 Permanent Technology Shocks
An important unresolved issue in macroeconomics is the extent to which perma-
nent innovations in technology can explain output fluctuations at the business cycle
frequency. While past research has claimed varying degrees of success, more re-
cent work has tempered enthusiasm for business cycle theories based on permanent
technology shocks. Cogley and Nason (1995) and Rotemberg and Woodford (1996)
demonstrate that the standard neoclassical model with permanent disembodied pro-
ductivity shocks is unable to match the persistence and comovement properties of
key aggregate variables. Christiano and Eichenbaum (1992) show that such models
predict excessive contemporaneous correlation between output growth and labor-
productivity growth. In this section, we extend this literature in two directions by
analyzing the effects of permanent embodied, as well as disembodied, technology
shocks and allowing for putty-clay technology. Greenwood et al. (1997) argue that
the evidence supports embodied technology as the primary source of technological
change, making the analysis of such shocks of particular interest. As shown below,
the putty-clay model generates dynamic responses to permanent technology shocks
that accord well with key properties of the data.
We begin with an analysis of the persistence properties of the two models. The
first row of table 1 shows the unconditional autocorrelation of output growth for the
two models for each type of technology shock.
17
For comparison, this statistic is es-
timated to be 0.3 in the data. An alternative measure of persistence, emphasized by
Rotemberg and Woodford (1996), is the ratio of standard deviations of forecastable
output growth to total output movements. The second and third rows of the table
report this statistic at the four- and eight-quarter horizons. For comparison, these
17
All model moments reported in this paper are computed using the linearized model.
21
Table 1: Output Persistence with Permanent Technology Shocks
Neoclassical Model Putty-clay Model
Disembodied Embodied Disembodied Embodied
cor(∆y
t
, ∆y
t−1
) 0.01 0.07 0.03 0.80
σ
∆ˆ y
t,4

∆y
t,4
0.06 0.25 0.14 0.83
σ
∆ˆ y
t,8

∆y
t,8
0.08 0.3 0.17 0.77
Notes: Shocks are permanent. y denotes the log of output. ∆ˆ y
t,j
=
E
t
(y
t+j
− y
t
), where expectations are based on date t information. All
reported moments are asymptotic means.
are estimated to be about 0.6–0.7 in the data.
Two results stand out clearly in table 1. First, the putty-clay model delivers
significantly more persistence in output growth (by either measure and for either
type of permanent technology shock) than the neoclassical model. Second, em-
bodied technology shocks generate much more persistence in output growth than
disembodied shocks. This is especially true for the putty-clay model, which, when
driven solely by permanent embodied technology shocks, actually overpredicts the
persistence in output growth observed in U.S. post war data.
The intuition behind these results is provided by the impulse responses to perma-
nent technology shocks, plotted in Figure 7. In both the putty-clay and neoclassical
models, the long run effect of a 1 − α percentage increase in technology is to raise
output, consumption and investment by 1% while leaving the long-run level of labor
unchanged. In the case of disembodied shocks (shown in the panels on the right)
total factor productivity increases immediately, causing an immediate expansion of
output. This initial increase in output represents a large fraction of the permanent
increase. As a result, for both the neoclassical and putty-clay models, nearly all
of the output dynamics are unforecastable, and output growth displays very little
persistence in response to disembodied shocks to technology.
If technology shocks are embodied in capital (shown in the panels on the left)
total factor productivity does not increase until the economy invests in new capital
goods. In the neoclassical model, a positive shock to embodied technology still
causes a large immediate expansion in output as hours surge in response to the high
22
Figure 7: Permanent Productivity Shocks
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 5 10 15 20
Output

years
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 5 10 15 20
Output

Putty-Clay
Neoclassical
years
-0.5
0.0
0.5
1.0
1.5
0 5 10 15 20

Consumption

years
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 5 10 15 20

Consumption

years
0.0
0.2
0.4
0.6
0.8
1.0
0 5 10 15 20

Labor

years
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0 5 10 15 20

Labor

years
Embodied Technology Shock Disembodied Technology Shock
rate of return to investment. Because of the rapid expansion in production, the
initial output response represents a large fraction of the permanent response and
output movements are mostly unpredictable. In the putty-clay model, a positive
shock to embodied technology causes only a small initial expansion despite the
increased desire for new investment, owing to the high costs of expanding production
in the short-run. Both output and hours expand slowly as new capital is brought
on-line, with labor reaching its peak response a number of years after the shock
occurs. As a result, most of the output dynamics are predictable and output growth
displays a high degree of persistence in response to embodied shocks to technology.
23
Table 2: Forecastable Comovement Between Output and Hours
Neoclassical Model Putty-clay Model
Disembodied Embodied Disembodied Embodied
cor(∆
ˆ
h
t,4
, ∆ˆ y
t,4
) -1.00 -1.00 0.44 0.44
cor(∆
ˆ
h
t,8
, ∆ˆ y
t,8
) -1.00 -1.00 0.28 0.28
Notes: h denotes the log of hours. See also Table 1.
In addition to the limited degree of internal propagation, Rotemberg and Wood-
ford (1996) criticize the high negative correlations between predictable movements in
output and hours implied by the standard neoclassical model in response to random
walk technology shocks. Table 2 formalizes this point by reporting the correlation
between forecastable growth of output and hours four and eight quarters ahead.
For comparison, these correlations are estimated to be about 0.86 in the data. For
the neoclassical model, the negative correlation follows from the fact that the peak
response in hours occurs at the onset of the shock while output continues to grow as
capital accumulation proceeds. Thus, while hours are falling in a predictable fash-
ion, output is rising in a predictable fashion. The putty-clay model’s slow expansion
of output and hours reverses this correlation and provides a closer match with the
data on this dimension.
Finally, we consider the models’ predictions regarding the contemporaneous cor-
relation between output and productivity growth. Table 3 reports this correlation
for the two models for each of the two sources of permanent productivity shocks.
For comparison, this correlation is estimated to be 0.76 in the data. If hours were
held constant, both models would predict a perfectly positive correlation between
growth in output and productivity. With disembodied shocks, this correlation is
nearly unity as the effect of the movements in hours on productivity are dwarfed by
the effects of the shock itself.
In the case of embodied technology shocks, the models’ predictions differ greatly.
The neoclassical model predicts a highly negative correlation between growth in
output and productivity. This negative correlation results from the immediate ex-
pansion of hours which produces a large decline in productivity at the same time as
the largest increase in output. The putty-clay model, on the other hand, predicts
24
Table 3: Output and Productivity Growth Comovement
Neoclassical Model Putty-clay Model
Disembodied Embodied Disembodied Embodied
cor(∆y
t
, ∆p
t
)) 0.99 -0.71 1.00 0.50
Notes: p denotes the log of output per hour. See also Table 1.
a positive correlation between output and productivity growth. This difference lies
in the muted expansion of hours, which does less to offset the positive comovement
in productivity and output directly resulting from the shock. The putty-clay model
thus provides an explanation for why the correlation between growth in output and
productivity may be positive but less than unity, even in the absence of shocks to
demand.
Although embodied shocks help explain a number of empirical regularities, some
results are not consistent with the business cycle. In particular, both the neoclas-
sical and putty-clay models create excess volatility of consumption and investment
relative to the data (as seen in the set of moments reported in the appendix). This
excess volatility occurs because factor cost movements create investment patterns
that overwhelm the usual desire to smooth consumption.
5 Estimation
The results in the previous section highlight the putty-clay model’s ability to explain
key business cycle facts such as persistence in output growth and positive predictable
comovement between output and hours. In this section we provide a more formal
evaluation of the empirical relevance of a putty-clay production process in matching
key moments of U.S. aggregate data.
To perform this evaluation, we construct a two-sector model that nests both the
putty-clay model and the neoclassical model within the same econometric frame-
work. Letting θ
t
denote the level of technology embodied in capital, we assume that
sector 1 output is derived from the putty-clay production process describe above,
while sector 2 output is derived from the Solow vintage-capital model described in
the appendix. Letting A
t
denote the level of disembodied technology, we assume
25
that final-goods output is a Cobb-Douglas function of sectoral output:
Y
t
= A
t
Y
λ
1t
Y
1−λ
2t
We then estimate λ, the share of output obtained from the putty-clay sector. If
our estimate of λ is close to unity, the data effectively put a large weight on putty-
clay production in order to match the vector of moments that we consider. If our
estimate of λ is close to zero, the data suggest little if any role for putty-clay in
explaining the moments we choose to match.
To address recent criticisms regarding standard RBC-style moment-matching ex-
ercises, our methodology relies on both unconditional moments traditionally empha-
sized in the RBC literature and conditional moments emphasized by Rotemberg and
Woodford (1996).
18
The unconditional moments include the standard deviations of
(the growth rates of) investment and hours relative to the standard deviation of
output, the correlations of investment and hours with output, and the first-order
autocorrelations of these variables. The conditional moments include correlations
and regression coefficients among predictable changes in output, investment and
hours over a four-quarter horizon.
19
These moments also include the ratio of the
standard deviation of the predictable change in output relative to the standard
deviation of the total change in output at this horizon.
To compute moments constructed from predictable components we specify a
VAR process for (y
t
, h
t
, i
t
), the logs of output, hours, and investment in the data.
20
As shown in the appendix, the statistical properties and hence all relevant moments
of this VAR are summarized by an unknown parameter vector ξ. We estimate ξ using
standard time-series techniques and then use the resulting parameter estimate to
compute a set of moments g(
ˆ
ξ), along with the variance of these moments V
g
.
Our estimation strategy is to choose λ, the share of output accounted for by the
18
Christiano and Eichenbaum (1992) provide a GMM procedure for estimating and evaluating
business-cycle models based on unconditional moments. A contribution of this paper is to provide
a GMM procedure that allows for both types of moments within a unified econometric framework.
19
Results using a combination of moments computed from predictable changes at the 4,8 and 16
quarter horizon do not alter our conclusions.
20
Our investment series is business fixed investment (non-residential equipment and structures).
In the model, investment is a linear combination of output and consumption so that consumption
and investment contain the same information when combined with output. An alternative approach
to the data is to define investment as a linear combination of output and non-durables consump-
tion. Although less desirable for a model explicitly designed to capture short-run capital/labor
complementarities we have considered this approach, as well as matching moments computed from
total consumption rather than investment. Neither of these alternatives alter our empirical results
in any substantial way.
26
putty-clay sector, along with other relevant parameters, to minimize the distance
between model moments and data moments. For a given vector ψ of unknown model
parameters, we use our model solution to compute g
M
(ψ), the model’s analog of g(ξ).
By minimizing
L(ψ) = (g
M
(ψ) −g(
ˆ
ξ))

V
−1
g
(g
M
(ψ) −g(
ˆ
ξ))
with respect to ψ we obtain the minimum distance estimator
ˆ
ψ. For a time-series
sample of size T, T ∗ L(
ˆ
ψ) provides a χ
2
test for equality between g
M
(
ˆ
ψ) and g(
ˆ
ξ).
For our moment matching exercise, we consider two independant sources of fluc-
tuations: disembodied technological change and embodied technological change. We
assume that shocks follow an AR1 process and then freely estimate [ρ
A
, ρ
θ
], the auto-
correlations of the shock processes, and
σ
θ
σ
A

θ
, the relative importance of embodied
shocks. To generalize our results beyond technology shocks, we also consider a model
that includes labor cost shocks. Under this specification, we also estimate the auto-
correlation of labor cost shocks, and the percentage of fluctuations attributable to
labor cost shocks.
Estimation results based on unconditional moments reported in table 4 place a
large weight on the putty-clay production technology – on the order of 50% for the
model that does not include labor cost shocks.
21
The estimate of
σ
θ
σ
A

θ
is about 0.2,
suggesting a substantial role for embodied technology shocks in explaining aggregate
dynamics.
22
For comparison purposes, table 4 also reports h
M
(ψ) and T ∗ L(ψ) for
the standard RBC model with disembodied shocks and ρ
A
= 0.95. Relative to this
baseline, allowing for embodied shocks and a non-zero weight on putty-clay provides
a substantial gain in terms of fit, reducing T ∗L(ψ) by 50%.
23
. In this specification,
the estimated values of ρ
A
and ρ
θ
reach their upper bound of unity, emphasizing
the importance of persistent shocks when matching unconditional moments.
Introducing labor cost shocks provides further gains in fit and places even greater
emphasis on putty-clay technology. The gain in fit comes from a relatively large
fraction of fluctuations being accounted for by labor cost shocks. Our estimation
results also set ρ
L
, the autocorrelation of labor cost shocks, at an imposed upper
21
The putty-clay share is estimated precisely with a standard error of about 0.02.
22
As an alternative to estimating
σ
θ
σ
A

θ
, we considered fixing this ratio at 0.6, Greenwood et
al. (1997)’s estimate of the share of post-war technological change embodied in capital. Assuming
random walk shocks and setting σL = 0, we estimate a putty-clay share of 0.69, indicating that our
estimate of the putty-clay share is robust to changes in the mix of technology shocks.
23
Nonetheless, we still reject a test of equality between model and data moments. This may
partly reflect the poor small-sample properties of such tests (Burnside and Eichenbaum 1996).
27
Table 4: Unconditional Moments: Estimation of Two-Sector Model
Neoclassical Two-Sector Data
Benchmark Model Est. S. E.
Model Parameters:
Putty-Clay Share 0 0.47 0.67
σ
θ
σ
θ

A
1
0 0.23 0.21
σ
L
σ
θ

A

L
2
0 0 0.55
ρ
θ
1.00 1.00
ρ
A
0.95 1.00 1.00
ρ
L
0.99
3
Moments:
σ
∆h

∆y
0.57 0.39 0.66 0.65 0.04
σ
∆i

∆y
2.85 2.49 2.39 1.80 0.11
cor(∆h
t
, ∆y
t
) 0.99 0.69 0.61 0.74 0.04
cor(∆i
t
, ∆y
t
) 1.00 0.87 0.89 0.61 0.05
cor(∆y
t
, ∆y
t−1
) -0.02 0.08 0.12 0.32 0.08
cor(∆h
t
, ∆h
t−1
) -0.04 0.21 0.26 0.61 0.06
cor(∆i
t
, ∆i
t−1
) -0.03 0.07 0.12 0.50 0.07
Minimized Objective: 708 361 246
1.
σ
θ
σ
θ
+σA
measures the share of technology shocks embodied in capital.
2.
σL
σ
θ
+σA+σL
measures the relative magnitude of labor market shocks.
3. Estimate constrained by upper bound of 0.99.
bound of 0.99. Thus, even with the introduction of labor cost shocks, our moment
matching exercise still places a strong emphasis on highly persistent shocks. Given
the importance of all three shocks in the estimation procedure, it is interesting
to ask what fraction of output variance is accounted for by each shock. Variance
decompositions using estimated parameter values imply that disembodied, embodied
and labor cost shocks account for 76%, 4% and 20% of output fluctuations at the
one-year horizon and 63%, 11% and 26% at the five-year horizon.
Estimation results based on the full set of moments reported in Table 5 also
imply a large putty-clay share – on the order of two-thirds, regardless of the shock
processes. Introducing putty-clay technology provides substantial gain in fit, which
is further improved through the introduction of labor cost shocks. When considering
the full set of moments, the estimate of
σ
θ
σ
A

θ
drops from about 0.2 to zero, and
the persistence of disembodied shocks falls to 0.94. The drop in
σ
θ
σ
A

θ
is offset
28
Table 5: All Moments: Estimation of Two-Sector Model
Neoclassical Two-Sector Data
Benchmark Model Est. S. E.
Model Parameters:
Putty-Clay Share 0 0.68 0.64
σ
θ
σ
θ

A
0 0.01 0.00
σ
L
σ
θ

A

L
0 0 0.64
ρ
θ
0.98
ρ
A
0.95 0.95 0.94
ρ
L
0.99
1
Moments:
σ
∆h

∆y
0.57 0.30 0.67 0.65 0.04
σ
∆i

∆y
2.85 2.60 2.67 1.80 0.11
cor(∆h
t
, ∆y
t
) 0.99 0.96 0.75 0.74 0.04
cor(∆i
t
, ∆y
t
) 1.00 0.99 0.99 0.61 0.05
cor(∆y
t
, ∆y
t−1
) -0.02 0.06 0.09 0.32 0.08
cor(∆h
t
, ∆h
t−1
) -0.04 0.24 0.25 0.61 0.06
cor(∆i
t
, ∆i
t−1
) -0.03 0.04 0.08 0.50 0.07
cor(∆
ˆ
h
t,4
, ∆ˆ y
t,4
) 0.89 0.90 0.82 0.89 0.04
cor(∆
ˆ
i
t,4
, ∆ˆ y
t,4
) 0.94 0.84 0.80 0.86 0.05
α
(∆
ˆ
h
t,4
,∆ˆ y
t,4
)
0.80 0.43 0.53 0.64 0.10
α
(∆
ˆ
i
t,4
,∆ˆ y
t,4
)
3.60 2.99 2.80 1.83 0.27
σ
∆ˆ y
t,4

∆y
t,4
0.27 0.25 0.26 0.62 0.07
Minimized Objective: 1148 786 444
1. Estimate constrained by upper bound of 0.99.
by an increase in the percentage of fluctuations obtained from labor cost shocks.
Computing variance decompositions using estimates in Table 5, we find that labor
cost shocks account for 15% of output fluctuations at the one-year horizon, and 47%
at the five-year horizon. These results imply that disembodied shocks to technology
do well at explaining output movements at high frequencies, while persistent shocks
to factor costs do better at lower frequencies.
24
24
As robustness checks to our estimation results, we have considered a number of issues, including
alternative parameterizations of the utility function and the presence of convex adjustment costs for
investment. For estimation based on unconditional moments, lowering the intertemporal elasticity
of substitution or adding adjustment costs increases both the putty-clay share and
σ
θ
σ
θ

A
. By
raising γ or introducing adjustment costs, we reduce the volatility of consumption and investment
29
6 Conclusion
By combining investment irreversibilities, capacity constraints, and variable capacity
utilization, the putty-clay model developed in this paper provides a rich framework
for analyzing a number of issues regarding investment, labor, capacity utilization,
and productivity. In this paper we highlight some implications for employment, out-
put, and investment at business cycle and medium-run frequencies. Compared to
standard neoclassical models, the putty-clay model displays a substantial degree of
persistence and propagation for both output and hours in response to shocks to fac-
tor costs and technology. And, unlike standard neoclassical models, the putty-clay
model generates forecastable comovements between labor, output, and consumption
consistent with the data. Finally, owing to the existence of a nonlinear aggregate
supply curve, the putty-clay model generates interesting asymmetries with reces-
sions steeper and deeper than expansions.
Beyond its descriptive appeal, the putty-clay production process is also found
to be empirically relevant for explaining business-cycle and medium-run dynamics.
Estimates obtained from a two-sector model that minimize the distance between
moments generated by the model and those obtained from the data place a sizable
weight on putty-clay production – on the order of one-half to three-fourths. This
finding is robust to the specification of the shock processes and the choice of mo-
ments used in the estimation. In addition to supporting a major role for putty-clay
technology, our results suggest that factor price shocks may be key to explaining
fluctuations, especially at the medium-run frequencies of two to eight years.
This paper focuses on the effects of factor-cost shocks and technological change
for business-cycle dynamics. The putty-clay model developed here has broader ap-
plications for fiscal, monetary and trade policy, and the study of transitional dynam-
ics for growing economies. In particular, this paper highlights the notion that the
short-run effects of policy may be substantially different from their medium-term
consequences, owing to the linkage between the capital accumulation process and
labor market dynamics.
in response to movements in factor costs. When considering the full set of moments we find little
sensitivity to such specifications.
30
References
Anderson, Gary and George Moore, “A Linear Algebraic Procedure for Solving
Linear Perfect Foresight Models,” Economics Letters, 1985, 17, 247–252.
Ando, Albert, Franco Modigliani, Robert Rasche, and Stephen
Turnovsky, “On the Role of Expectations of Price and Technological Change
in an Investment Function,” International Economic Review, 1974, 15, 384–414.
Atkeson, Andrew and Patrick J. Kehoe, “Putty-Clay Capital and Energy,”
August 1994. NBER Working Paper No. 4833.
Basu, Susanto and John G. Fernald, “Returns to Scale in U.S. Production:
Estimates and Implications,” Journal of Political Economy, 1997, 105, 249–83.
Benhabib, Jess and Aldo Rustichini, “Vintage Capital, Investment and
Growth,” Journal of Economic Theory, 1991, 55, 323–339.
and , “A Vintage Capital Model of Investment and Growth: Theory and
Evidence,” in Becker R. et al., ed., General Equilibrium, Growth and Trade. II
The Legacy of Lionel W. McKenzie, New York: Academic Press, 1993, pp. 248–
301.
Bischoff, Charles W., “The Effect of Alternative Lag Distributions,” in
G. Fromm, ed., Tax Incentives and Capital Spending, Washington, D.C.: The
Brookings Institution, 1971, chapter 3, pp. 61–130.
Blanchard, Olivier J., “The Medium Run,” Brookings Papers on Economic Ac-
tivity, 1997, pp. 89–158.
Boucekkine, Raouf, Marc Germain, and Omar Licandro, “Replacement
Echoes in the Vintage Capital Growth Model,” Journal of Economic Theory,
1997, 74, 333–48.
Bresnahan, Timothy F. and Valerie A. Ramey, “Output Fluctuations at the
Plant Level,” Quarterly Journal of Economics, 1994, 109, 593–624.
Burnside, Craig and Martin Eichenbaum, “Small Sample Properties of GMM
Based Wald Tests,” Journal of Business Economics and Statistics, 1996.
31
Caballero, Ricardo J. and Mohamad L. Hammour, “On the Timing and
Efficiency of Creative Destruction,” Quarterly Journal of Economics, 1996,
111, 805–52.
and , “Incomplete Contracts, Factor Proportions, and Unemployment,”
Carnegie Rochester Conference Series on Public Policy, 1998, 48.
, Eduardo Engel, and John Haltiwanger, “Plant-Level Adjustment and
Aggregate Investment Dynamics,” Brookings Papers on Economic Activity,
1995, pp. 1–54.
Calvo, Guillermo A., “Optimal Growth in a Putty-Clay Model,” Econometrica,
1976, 44, 867–878.
Campbell, Jeffrey R., “Technical Change, Diffusion and Productivity,” 1994.
Mimeo, Rochester University.
Cass, David and Joseph E. Stiglitz, “The Implications of Alternative Saving and
Expectations Hypotheses for Choices of Technique and Patterns of Growth,”
Journal of Political Economy, 1969, 77, 586–627.
Christiano, Lawrence J. and Martin Eichenbaum, “Current Real-Business-
Cycle Theories and Aggregate Labor-Market Fluctuations,” American Eco-
nomic Review, 1992, 82, 430–450.
Cogley, Timothy and James M. Nason, “Output Dynamics in Real Business
Cycle Models,” American Economic Review, 1995, 85, 492–511.
Cooley, Thomas F., Gary D. Hansen, and Edward C. Prescott, “Equi-
librium business cycles with idle resources and variable capacity utilization,”
Economic Theory, 1995, 6, 35–49.
, Jeremy Greenwood, and Mehmet Yorukoglu, “The Replacement Prob-
lem,” 1994. Institute for Empirical Macroeconomics Discussion Paper 95.
Cooper, Russel and John Haltiwanger, “The Macroeconomic Implications of
Machine Replacement: Theory and Evidence,” American Economic Review,
1993, 83, 360–382.
32
, , and Laura Power, “Machine Replacement and the Business Cycle:
Lumps and Bumps,” 1995. NBER Working Paper #5260.
Cover, James Perry, “Asymmetric Effects of Positive and Negative Money-Supply
Shocks,” Quarterly Journal of Economics, 1992, 107, 1261–1282.
Doms, Mark and Timothy Dunne, “An Investigation into Capital and Labor
Adjustment at the Plant Level,” 1993. Mimeo, Center for Economic Studies,
Bureau of the Census.
Fair, Ray C. and John B. Taylor, “Solution and Maximum Likelihood Esti-
mation of Dynamic Nonlinear Rational Expectations Models,” Econometrica,
July 1983, 51 (4), 1169–1185.
Gilchrist, Simon and John C. Williams, “Cutting Models Down to Size: Effi-
cient Methods for Solving Large Rational Expectations Models,” 1998. Mimeo,
Board of Governors of the Federal Reserve.
and , “Putty-Clay and Investment: A Steady State Analysis,” 1998.
Mimeo, Board of Governors of the Federal Reserve.
Greenwood, Jeremy, Zvi Hercowitz, and Per Krusell, “Long-Run Implica-
tions of Investment-Specific Technological Change,” American Economic Re-
view, 1997, pp. 342–362.
Hamilton, James D., Time Series Analysis, Princeton University Press, 1994.
Hansen, Gary, “Indivisible Labor and the Business Cycle,” Journal of Monetary
Economics, 1985, 16, 309–327.
Johansen, Leif, “Substitution versus Fixed Production Coefficients in the Theory
of Economic Growth: A Synthesis,” Econometrica, 1959, 27, 157–176.
Johnson, Norman L., Samuel Kotz, and N. Balakrishnan, Continuous Uni-
variate Distributions, Volume 1, Second Edition, New York: John Wiley &
Sons, 1994.
Judd, Kenneth L., “Projection Methods for Solving Aggregate Growth Models,”
Journal of Economic Theory, December 1992, 58, 410–452.
, Numerical Methods in Economics, Cambridge, MA.: MIT Press, 1998.
33
Kydland, Finn E. and Edward C. Prescott, “Time to Build and Aggregate
Fluctuations,” Econometrica, 1982, 50, 1345–1370.
and , “Hours and Employment Variation in Business Cycle Theory,”
Economic Theory, 1991, 1 (1), 63–81.
Lucas, Robert E., “Understanding Business Cycles,” Carnegie Rochester Confer-
ence Series on Public Policy, 1977, 5, 7–29.
Malinvaud, Edmond, Profitability and Unemployment, Cambridge, England.:
Cambridge Press, 1980.
Mork, Knut A., “Oil and the Macroeconomy when Prices go Up and Down,”
Journal of Political Economy, 1989, 79, 173–189.
Neftci, Salih N., “Are Economic Time Series Asymmetric over the Business Cy-
cle?,” Journal of Political Economy, 1984, 92, 307–28.
Phelps, Edmund S., “Substitution, Fixed Proportions, Growth and Distribution,”
International Economic Review, 1963, 4, 265–288.
Potter, Simon M., “A Nonlinear Approach to U.S. GNP,” Journal of Applied
Econometrics, 1995, 10, 109–125.
Rogerson, Richard, “Indivisible Labor, Lotteries and Equilibrium,” Journal of
Monetary Economics, 1988, 21, 3–16.
Rotemberg, Julio J. and Michael Woodford, “Real Business-Cycle Models and
the Forecastable Movements in Output, Hours, and Consumption,” American
Economic Review, 1996, 86, 71–89.
Sheshinski, E., “Balanced Growth and Stability in the Johansen Vintage Model,”
Review of Economic Studies, 1967, 34, 239–248.
Sichel, Daniel E., “Business Cycle Asymmetry: A Deeper Look,” Economic In-
quiry, 1993, 31, 224–236.
Solow, R.M., “Substitution and Fixed Proportions in the Theory of Capital,”
Review of Economic Studies, 1962, 29, 207–218.
Tatom, John A., “Are the Macroeconomic Effects of Oil Changes Symmetric?,”
Carnegie Rochester Conference Series on Public Policy, 1988, 28, 325–368.
34
Appendix
In this appendix we describe the derivation of the neoclassical vintage model, the
solution methods for the dynamic model, details of the econometric procedure along
with a complete description of the two-sector model used in section 5. At the end of
this appendix we also provide complete tables for the moments of the neoclassical
and putty-clay models with permanent technology shocks.
Derivation of the Neoclassical Vintage Model:
For the purpose of comparison, we construct a neoclassical model of vintage capital,
initially introduced by Solow (1962). In this model, the restriction that ex post
capital-labor ratios are fixed is removed. The two models are otherwise identical.
Let I
t−j
denote aggregate investment in period t−j, and define the following capital
aggregator
K
t
=
M
¸
j=1
θ
1/α
t−j
(1 −δ
j
)I
t−j
Period t labor and output of machine i created in period t −j satisfies:
L
t,t−j
(i) = ((1 −α)θ
it−j
/W
t
)
1/α
(1 −δ
j
)I
t−j
and
Y
t,t−j
(i) = θ
i,t−j
1/α
((1 −α)/W
t
)
(1−α)/α
(1 −δ
j
)I
t−j
where (1 −δ
j
)I
t−j
represents the capital remaining at time t that was put in place
at time t − j. Integrating over machines with respect to the distribution of θ
i,t−j
,
and using the result that
E(θ
i,t−j
1/α
) = θ
1/α
t−j
exp(
σ
2

1 −α
α
)
provides the relationship between total vintage t −j output and labor inputs:
(1 −α)Y
t,t−j
= W
t
L
t,t−j
where
Y
t,t−j
= A
1−α

θ
1/α
t−j
(1 −δ
j
)I
t−j

α
L
1−α
t,t−j
and A = exp(
σ
2

). Summing across vintages we obtain
Y
t
= ((1 −α)A/W
t
)
(1−α)/α
K
t
35
L
t
= ((1 −α)A/W
t
)
1/α
K
t
.
Combining these two expressions gives the aggregate production function
Y
t
= A
1−α
K
α
t
L
1−α
t
where L
t
is aggregate labor input and A = exp(
σ
2

) is a scale correction that results
from aggregating across machines at the idiosyncratic level.
If we assume that δ
j
= 1 − (1 − δ)
j−1
and M = ∞, we obtain the following
capital accumulation equation
K
t
= (1 −δ)K
t−1

1/α
t−1
I
t−1
Thus, shocks to embodied technological change are identical to shocks to the true
economic cost of new capital goods.
Solution Methods:
The model consists of 2M + N state variables, including average machine efficien-
cies, X
t−j
, j = 1, . . . , M, and the quantity of new machines per vintage, Q
t−j
, j =
1, . . . , M, for each of the M vintages in existence, and N shocks. Owing to the
log-normal distribution of X
it
, these state variables completely summarize the ex-
isting distribution of machines depicted in Figure 1. By choosing M sufficiently
large we provide an arbitrarily good approximation to the case M = ∞. For values
of M large enough to analyze business-cycle frequency properties of the model, the
state-space is too large for the type of nonlinear state-space methods discussed in
Judd (1998). Instead, depending on the purpose, we apply one of two methods that
yield approximate solutions at relatively low computational cost.
The first method uses a log-linearization around the deterministic steady state of
the equations describing the equilibrium. The resulting linear model, which includes
M leads and lags of ln(X) and ln(Q), is then solved using the AIM implementation
of the Blanchard-Kahn method due to Anderson and Moore (1985). This algorithm
yields accurate solutions for values at relatively low computational cost for M up to
40, which is sufficiently large for an annual version of the model. For the quarterly
version of the model, M = 40 implies capital goods completely depreciate within 10
years, which seems to be an unrealistically short lifespan. Thus, following Gilchrist
and Williams (1998a), we use polynomial distributed leads and lags to approximate
the M = ∞ leads and lags of variables. For example, for some variable u and
36
coefficient sequence {a
j
}

1
, we approximate the sum
¸

j=1
a
j
u
t−j
by the polynomial
distributed lag (or lead)
u
t−1
B(L)
, where B(L) = b
0
−b
1
L−b
2
L
2
−. . .−b
p
L
p
and p is finite.
For the putty-clay model we use p = 1 and chose values corresponding to b
0
and
b
1
that minimize the weighted squared deviations between the PDL representation
and the original lag or lead structure; this approximations yields virtually no loss
in accuracy for model simulations. This adds 8 equations to our model and reduces
the maximum lead and lag from M to 1, thus drastically reducing the size of the
companion form of the model. The solution time for this approximate model is
trivial.
The log-linearized model, approximated in the manner described above, is used
for the simulations reported in the paper except for those illustrating the asym-
metrical response to large shocks shown in figure 6, which requires a method that
preserves the nonlinearities of the model. For those simulations an extended path
algorithm based on Fair and Taylor (1983) is used for a model with M = 160. This
method requires far more memory and CPU time than the linearization method
described above. For small shocks, the two methods yield nearly identical answers.
Both of these methods compute “certainty-equivalent” solutions; that is, expec-
tations are computed assuming all future shocks equal their mean values of zero. To
address this issue, we used projection methods to solve a version of the model with
small M. This approach, described in Judd (1992), uses polynomial approxima-
tions to the decision rules and multi-point quadrature to approximate expectation
integrals and therefore does not impose certainty equivalence on the solutions. We
found that the solutions were very close to those generated using the extended
path algorithm, implying that the certainty equivalent solution provides a good ap-
proximation even in the presence of substantial non-linearities and reasonably large
aggregate shocks.
Econometric Methodology for Moment Matching Exercise
In this section of the paper we present our moment matching methodology. We start
by specifying a stochastic process for the log of output, hours and investment in the
data which depends on an unknown parameter vector θ which is to be estimated.
Our approach follows Rotemberg and Woodford’s method of specifying a tightly
parameterized low-order VAR system to characterize the data. Our data consists
of private non-farm output, total private hours and non-residential business fixed
37
investment (equipment and structures) for the period 1960-1997. Relative to GDP,
the output and hours series exclude government and farm output as well as the
imputed output obtained from owner-occupied housing. The output and hours
series are thus defined in a mutually consistent manner. With the exception of
agricultural investment in structures and machinery, the investment series is also
consistent with the output and hours series. Both the output and the investment
series are deflated using 1997 chain weighted-deflators.
After first removing a linear time trend from the hours series, we assume that
the first difference of the log of output, the log of the investment/output ratio and
the log of hours, [∆y
t
, i
t
−y
t
, h
t
] can be represented using a two-lag stationary VAR
representation. Defining
u

t
= [∆y
t
, i
t
−y
t
, h
t
, ∆y
t−1
, i
t−1
−y
t−1
, h
t−1
], E(u
t
u

t
) = Σ
u
e

t
= [e
y
t
, e
i
t
, e
h
t
], E(e
t
e

t
) = Σ
e
, E(e
t
e

t+s
) = 0 for s = 0
we express the stochastic process for u
t
in companion form as:
u
t
= Au
t−1
+v
t
, A =
¸
Π
I 0
¸
, v
t
=
¸
e
t
0
¸
, E(v
t
v

t
) = Σ
v
where Π is a matrix of VAR coefficients. Defining D
n
as the duplication matrix such
that D
n
vech(Σ
e
) = vec(Σ
e
), the parameter vector ξ and its associated variance is
then:
ξ =
¸
vec(Π)
vech(Σ
e
)
¸
, V
ξ
=
¸
Σ
e
⊗Σ
−1
u
0
0 Σ
22
¸
where Σ
22
= 2D
+
n

e
⊗ Σ
e
)(D
+
n
)

for D
+
n
= (D

n
D
n
)
−1
D

n
. (Hamilton (1994) pp
301-302 provides details.)
Given this specification for the stochastic process for output, hours and invest-
ment, we are interested in computing second moments of both the kth differences
of these variables as well as second moments of the forecastable components of the
kth differences, where the forecast is made conditional on time t − k information.
To obtain expressions for these second moments as functions of the underlying VAR
parameters, we first express the kth differences in y
t+k
, i
t+k
and h
t+k
as functions
of data known at time t and shocks that occur between t and t + k. For x = y, i, h
we have:
∆x
t,k
≡ x
t+k
−x
t
= b
x
k
u
t
+
k
¸
j=1
d
x
j
v
t+j
38
where
b
y
k
= e

1
k
¸
i=1
A
i
, b
i
k
= b
y
k
+e

2
(A
k
−I), b
h
k
= e

3
(A
k
−I)
d
y
j
= e

1
j
¸
s=1
A
s−1
, d
i
j
= d
y
j
+e

2
A
j−1
, d
h
j
= e

3
A
j−1
.
Taking expectations as of time t, the predictable components of the kth difference
of x is:
∆ˆ x
t,k
≡ E
t
{x
t+k
−x
t
} = b
x
k
u
t
Computing second moments, we obtain an expressions for the variance of ∆x
t,k
σ
2
x,k
≡ E(∆x
t,k
)
2
= (b
x
k

u
(b
x
k
)

+
k
¸
j=1
(d
x
j

v
(d
x
j
)

The covariance between ∆x
t,k
, and ∆y
t,k
is obtained from
σ
xy,k
≡ E(∆x
t,k
∆y
t,k
) = (b
x
k

u
(b
y
k
)

+
k
¸
j=1
(d
x
j

v
(d
y
j
)

.
Similarly, the variance of the predictable component of the kth difference in x is
simply
σ
2
ˆ x,k
≡ E(∆ˆ x
t,k
)
2
= (b
x
k

u
(b
x
k
)

while the covariance between the predictable components of the kth differences of x
with y is obtained from:
σ
ˆ xy,k
≡ E(∆ˆ x
t,k
∆ˆ y
t,k
) = (b
x
k

u
(b
y
k
)

To compute autocorrelations among kth differences, note that E(∆x
t+k,k
∆x
t,k
) =
E(∆x
t,2k
∆x
t,k
) − E(∆x
t,k
)
2
. From the expressions obtained above we then have
E(∆x
t,2k
∆x
t,k
) −E(∆x
t,k
)
2
= b
x
2k
Σ
u
(b
x
k
)

−b
x
k
Σ
u
(b
x
k
)

implying that
ρ
x,k

E(∆x
t+k,k
, ∆x
t,k
)
E(∆x
t,k
)
2
=
(b
x
2k
−b
x
k

u
(b
x
k
)

(b
x
k

u
(b
x
k
)

Using these expressions, we define two sets of moments. The first set of moments
are the unconditional standard deviations of ∆i
t,k
, ∆h
t,k
relative to the uncondi-
tional standard deviation of ∆y
t,k
, the unconditional correlations of ∆i
t,k
, ∆h
t,k
with ∆y
t,k
, and the autocorrelations of ∆y
t,k
, ∆i
t,k
and ∆h
t,k
:
h
1,k
=
¸
σ
i,k
σ
y,k
,
σ
h,k
σ
y,k
,
σ
iy,k

i,k
σ
y,k
)
,
σ
hy,k

h,k
σ
y,k
)
, ρ
y,k
, ρ
i,k
, ρ
h,k
¸

.
39
For k = 1 these moments are the unconditional moments considered in section 6.
The second set of moments computes regression coefficients and correlations
between the predictable components of ∆i
t,k
, ∆h
t,k
and the predictable component
of ∆y
t,k
. According to Rotemberg and Woodford, these measures of the underlying
dynamic response of the system when away from steady state. Intuitively, these
statistics capture the strength and magnitude of the expected comovements among
output, investment and hours. By varying k, we vary the horizon over which the
comovements between forecastable components are computed. The second set of
moments also includes the ratio of the variance of the predictable component of
∆y
t,k
relative to the total variance of ∆y
t,k
. Again, Rotemberg and Woodford view
this statistic as providing a good indicator of the degree to which the model contains
a propagation mechanism. Thus the second set of moments may be written as:
g
2,k
=
¸
σ
ˆ
iy,k
σ
2
ˆ y,k
,
σ
ˆ
hy,k
σ
2
ˆ y,k
,
σ
ˆ
iy,k
σ
ˆı,k
σ
ˆ y,k
,
σ
ˆ
hy,k
σ
ˆ
h,k
σ
ˆ y,k
,
σ
ˆ y,k
σ
y,k
¸

.
where the first two elements are regression coefficients, the second two elements
are correlations, and the final element is the variance ratio. Recognizing that both
g
1,1
and g
2,k
are functions of A, Σ
u
, Σ
v
and hence ultimately functions of the VAR
parameter vector ξ, we obtain our moment vector of interest: g(ξ) = [g

1,1
g

2,k
]

. To
obtain an estimate of the variance of this moment vector, we use a Taylor series
expansion of g (the delta method) to obtain V
g
=
∂g
∂ξ

V
ξ
∂g
∂ξ
.
We estimate the model parameters ψ = [a, ρ
A
, ρ
θ
, ρ
L
,
σ
θ
σ
θ

A
,
σ
L
σ
θ

A

L
] by min-
imizing the distance between g(ξ) and g
M
(ψ) where g
M
is the model’s analog of
g(ξ). To obtain g
M
as a function of ψ we rely on the fact that our model solution
is linear and may be expressed in the first order companion form:
X
t
= AX
t−1
+BU
t
where E(U
t
U

t
) = I and X
t
is a vector of model variables with y
t
= e

y
X
t
, c
t
= e

y
X
t
and h
t
= e

h
X
t
. The moment vector g
M
may then be computed as a function of
A, B. Because the matrices A, B are (nonlinear) functions of the underlying model
parameters ψ, the moment vector g
M
is also a function of ψ. Standard errors for ψ
may then be obtained from V
ψ
= {
∂g
M
∂ψ

V
−1
g
∂g
M
∂ψ
}
−1
.
Two Sector Model:
In this section we provide a full description of the system of equations that char-
acterize the equilibrium of the two-sector model. We set the final-goods price as
40
the numeraire and define P
1,t
and P
2,t
as the relative output prices for each sector.
Disembodied technology, A
t
, affects both sectors equally.
25
We define η
t
as the labor
market shock.
Sectoral Production:
Y
1,t
=
M
¸
j=1

1 −Φ(z
t−j
t
−σ)

(1 −δ
j
)Q
t−j
X
t−j
L
1,t
=
M
¸
j=1

1 −Φ(z
t−j
t
)

(1 −δ
j
)Q
t−j
Y
2,t
= K
α
t
L
1−α
2,t
K
t
= (1 −δ)K
t−1

1/α
t−1
I
t−1
.
Optimality conditions for sectoral labor and capital accumulation:
k
t
= αE
t

M
¸
j=1
˜
R
t,t+j
(1 −δ
j
)

1 −Φ(z
t
t+j
−σ)

P
1,t+j
X
t
¸
k
t
= E
t

M
¸
j=1
˜
R
t,t+j
(1 −δ
j
)

1 −Φ(z
t
t+j
−σ)

P
1,t+j
X
t

M
¸
j=1
˜
R
t,t+j
(1 −δ
j
)

1 −Φ(z
t
t+j
)

W
t+j
¸
θ
−1/α
t
= E
t
R
t,t+1

αP
2,t+1
Y
2,t+1
K
t+1
+ (1 −δ)θ
−1/α
t+1

(1 −α)P
2,t
Y
2,t
L
2,t
= W
t
φ(z
t−j
t
−σ)P
1,t
X
t−j
φ(z
t−j
t
)
= W
t
, j = 1, ...M.
This last expression equates marginal products across all vintages. It could have
alternatively been written as z
t−j
t

1
σ

log W
t
−log P
1,t
X
t−j
+
1
2
σ
2

.
Household first order conditions:
U
c,t
=
β
1 +n
E
t
R
t,t+1
U
c,t+1
25
Note that the effect of the level of disembodied technology on the first-order conditions of firms
is captured by the output price terms.
41
U
c,t
η
t
W
t
+U
L,t
= 0.
Aggregate output and resource constraints:
Y
t
= A
t
Y
λ
1,t
Y
1−λ
2,t
C
t
= Y
t
−k
t
Q
t
−I
t
L
t
= L
1,t
+L
2,t
.
Shocks:
ln(A
t
) = ρ
A
ln(A
t−1
) +e
A,t
, with E(e
2
A,t
) = σ
2
A
ln(θ
t
) = ρ
θ
ln(θ
t−1
) +e
θ,t
, with E(e
2
θ,t
) = σ
2
θ
ln(η
t
) = ρ
L
ln(η
t−1
) +e
η,t
, with E(e
2
η,t
) = σ
2
L
Permanent Technology Shocks:
Tables 6 and 7 report the unconditional and forecastable moments, respectively, re-
sulting from permanent technology shocks for the two models as well as the statistics
estimated in the data.
42
Table 6: Unconditional Moments – Permanent Technology Shocks
Neoclassical Model Putty-Clay Model Data
Disemb. Embod. Disemb. Embod. Est. S. E.
σ
∆c

∆y
0.55 1.13 0.70 2.87 0.36 0.02
σ
∆h

∆y
0.36 1.57 0.09 0.91 0.66 0.04
σ
∆i

∆y
2.15 6.03 1.76 8.14 1.82 0.11
σ
∆p

∆y
0.65 0.69 0.91 0.77 0.68 0.04
cor(∆c
t
, ∆y
t
) 0.99 -0.83 0.99 -0.12 0.52 0.06
cor(∆h
t
, ∆y
t
) 0.98 0.95 0.96 0.68 0.74 0.04
cor(∆i
t
, ∆y
t
) 0.99 0.97 0.99 0.53 0.60 0.05
cor(∆p
t
, ∆y
t
) 0.99 -0.71 1.00 0.50 0.76 0.03
cor(∆y
t
, ∆y
t−1
) 0.01 0.07 0.04 0.80 0.31 0.08
cor(∆c
t
, ∆c
t−1
) 0.10 0.05 0.10 0.06 0.34 0.08
cor(∆h
t
, ∆h
t−1
) -0.03 -0.03 0.20 0.20 0.60 0.06
cor(∆i
t
, ∆i
t−1
) -0.02 -0.02 -0.01 -0.02 0.50 0.07
cor(∆p
t
, ∆p
t−1
) 0.06 0.18 0.03 0.73 0.60 0.06
Table 7: Forecastable Moments – Permanent Technology Shocks
Neoclassical Model Putty-Clay Model Data
Disemb. Embod. Disemb. Embod. Est. S. E.
cor(∆ˆ c
t,4
, ∆ˆ y
t,4
) 1.00 1.00 1.00 1.00 0.95 0.07
cor(∆
ˆ
h
t,4
, ∆ˆ y
t,4
) -1.00 -1.00 0.44 0.44 0.86 0.06
cor(∆
ˆ
i
t,4
, ∆ˆ y
t,4
) -1.00 -1.00 -0.98 -0.98 0.76 0.07
cor(∆ˆ p
t,4
, ∆ˆ y
t,4
) 1.00 1.00 0.95 0.95 0.76 0.10
α
(∆ˆ c
t,4
,∆ˆ y
t,4
)
3.18 3.18 1.61 1.61 0.22 0.07
α
(∆
ˆ
h
t,4
,∆ˆ y
t,4
)
-1.68 -1.68 0.14 0.14 0.59 0.09
α
(∆
ˆ
i
t,4
,∆ˆ y
t,4
)
-4.44 -4.44 -0.49 -0.49 1.70 0.24
α
(∆ˆ p
t,4
,∆ˆ y
t,4
)
2.68 2.68 0.86 0.86 0.41 0.09
σ
∆ˆ y
t,4

∆y
t,4
0.06 0.25 0.14 0.83 0.64 0.06
σ
∆ˆ y
t,8

∆y
t,8
0.08 0.30 0.17 0.77 0.71 0.07
σ
∆ˆ y
t,16

∆y
t,16
0.09 0.31 0.18 0.67 0.65 0.06
43